28 March 2024 in IEEE Internet of Things Journal
The Industrial Internet of Things (IIoT) has significantly advanced traditional industrial systems, especially in facilitating remote monitoring and predictive maintenance for computer numerical control (CNC) machines. Ball screw drives (BSDs), crucial in CNC machining, require regular upkeep, often challenged by environmental influences and the limitations of wired sensor-based diagnostic procedures. Wireless sensors offer a cost-effective solution but struggle with data integrity during transmission, impacting remote wear evaluation of BSDs. Current recovery methods are not always adequate, often relying on extensive historical data and suffering from accumulating errors. Addressing these limitations, a novel self-supervised transfer learning (SSTL) model is proposed for remote wear assessment of machine tool components. This model integrates an image capture module into CNC surveillance systems and employs a LocalMIM module, which is pretrained and fine-tuned to adapt to various domains, especially for image quality deterioration during data transmission. The SSTL model is designed to function effectively despite significant pixel data loss, eliminating the need for historical data for image restoration. This innovation is particularly adept at evaluating wear in environments with compromised image transmission, providing a robust predictive maintenance strategy for BSDs that is less affected by data loss, aiding a pressing industrial requirement for consistent and dependable remote monitoring solutions.
23 January 2025 in IEEE Transactions on Reliability
In real-world applications, the diagnostic efficiency of rolling bearings is commonly affected by operating conditions like fluctuating rotating speed and varying loads, especially, environmental disturbances like transient noises. These disturbances tend to mask the indicators of damage, presenting substantial obstacles for accurately pinpointing failures. Traditional diagnostic methods struggle with the complexity and the noise sensitivity of such scenarios, often failing to accurately identify failure signs amidst multivariate random transient noise. To address these challenges, the current study proposes a method known as short-term Markov transition frequency peak rate. This method focuses on precisely tracking temporal state changes and identifying abnormal signals. It is aimed at mitigating transient noise interference at its source and enhancing insensitivity to external transient noise, which facilitates a more accurate and reliable selection of demodulation bands. Furthermore, an amplitude interference-limiting mechanism is designed within this method to discern and mitigate the impact of transient noise that may adversely affect the demodulation band selection process. The experimental results validate the effectiveness of this approach, demonstrating that it can reliably diagnose bearing faults even in the presence of transient disturbances.
17 April 2023 in IEEE Transactions on Instrumentation and Measurement
X-ray imaging is a common non-destructive detection method for Carbon Fiber Composite Structures (CFCS) that is useful in identifying damage in CFCS-cored wires. In recent years, deep learning models that incorporate classification and objection detection have become frequently utilized by the non-destructive testing industry. These models typically rely on the assumption that there are sufficient annotated failure samples from history that have been measured and can be used for training. Unfortunately, in real-world measurements, it is often challenging to obtain these types of samples. To address the issue of small sample size in such scenarios of real-world field testing, this paper propose a mixed samples-driven methodology based on the Denoising Diffusion Probabilistic Model (DDPM) for identifying damage in CFCS. First, new samples are synthesized through DDPM module to improve the robustness of a small sample size. Then, the synthesized sample, along with a small number of authentic samples measured from real-world testing, are then integrated and fed into a DenseNet-based module. Lastly, the mixed samples-driven architecture is then constructed and employed to diagnose the damage of CFCS. The effectiveness of this approach is demonstrated through experiments in real-world field testing.
20 November 2024 in IEEE Transactions on Instrumentation and Measurement
Machine tool elements, including ball screw drives (BSDs) in CNC machines, play a critical role but are vulnerable to wear due to external factors. Existing approaches in the field of fault diagnosis and health condition monitoring involve the examination of factors such as speed, loads, and external influences in real-world operating conditions. Specifically, these deep learning methods face challenges associated with the requirement for extensive historical datasets for training, a constraint that may not be available in numerous practical scenarios. Furthermore, their performance heavily depends on robust data correlations. On the other hand, vision-oriented techniques, utilizing visual data from sensors, provide a more comprehensive understanding and an advantage in accurately detecting surface damage. Nevertheless, they encounter challenges such as contamination by oil or other contaminants. To address these challenges, the scale-aware domain adaptation (SADA) framework has been proposed. This methodology incorporates three designed modules: the focused linear attention modules (FLAM), the scale-aware feature stacking (SAFS) block, and the nonlinear activation-free (NAF) block. Initially, SADA is trained on an open-access source domain dataset. Utilizing the SAFS block allows for the prioritization of influential features during the learning process through weighted assessments. Meanwhile, FLAM is designed to refine the efficiency of attention mechanisms for data sequences of considerable length. The NAF block aids in effortlessly fusing features at various scales. Upon this foundational training, the weights from the Encoder and Decoder are generalized for application to the target domain, specifically datasets concerning BSDs. SADA’s effectiveness is evidenced through case studies in accurately detecting surface damage on BSDs, contributing to the improved durability of machine tool components.
1 April 2024 in IEEE Transactions on Instrumentation and Measurement
In real-world scenarios, the performance of rolling bearings can be significantly affected by frequent environmental disturbances, including high-energy fluctuations, transient noises, unintended collisions, or sudden variations in loading. These disturbances have the potential to obscure signs of damage, posing a challenge to the identification of failures. Traditional techniques for selecting demodulation band encounter challenges due to inherent complexity and parameter sensitivity. These methods, which heavily rely on post-processing signal analysis involving interference components, prove ill-suited for complex scenarios, especially faced with multivariate random pulse noise. To address these challenges, the paper proposes a method for Markov modeling of signal condition transitions to mitigate transient noise interference at its source. This method aims to improve insensitivity for external transient noise, enabling a more accurate and reliable selection of demodulation bands. The proposed method employs a signal transition model grounded in a Markov transition matrix, coupled with wavelet transforms and reconstruction. It improves the identification of anomalous signal components and leverages the transition matrix to track temporal state changes. Additionally, an amplitude interference-limiting mechanism is designed to identify and mitigating transient noise that may adversely affect the demodulation band selection process. The study’s results demonstrate that this mechanism allows typical Kurtogram techniques to accurately identify bearing fault-characteristic bands, and the experimental results validate the effectiveness of the proposed methodology for fault diagnosis of rolling bearings.